A $2 million contract that United States Immigration and Customs Enforcement signed with Israeli commercial spyware vendor Paragon Solutions has been paused and placed under compliance review, WIRED has learned.

The White House’s scrutiny of the contract marks the first test of the Biden administration’s executive order restricting the government’s use of spyware.

The one-year contract between Paragon’s US subsidiary in Chantilly, Virginia, and ICE’s Homeland Security Investigations (HSI) Division 3 was signed on September 27 and first reported by WIRED on October 1. A few days later, on October 8, HSI issued a stop-work order for the award “to review and verify compliance with Executive Order 14093,” a Department of Homeland Security spokesperson tells WIRED.

The executive order signed by President Joe Biden in March 2023 aims to restrict the US government’s use of commercial spyware technology while promoting its “responsible use” that aligns with the protection of human rights.

DHS did not confirm whether the contract, which says it covers a “fully configured proprietary solution including license, hardware, warranty, maintenance, and training,” includes the deployment of Paragon’s flagship product, Graphite, a powerful spyware tool that reportedly extracts data primarily from cloud backups.

“We immediately engaged the leadership at DHS and worked very collaboratively together to understand exactly what was put in place, what the scope of this contract was, and whether or not it adhered to the procedures and requirements of the executive order,” a senior US administration official with first-hand knowledge of the workings of the executive order tells WIRED. The official requested anonymity to speak candidly about the White House’s review of the ICE contract.

Paragon Solutions did not respond to WIRED's request to comment on the contract's review.

The process laid out in the executive order requires a robust review of the due diligence regarding both the vendor and the tool, to see whether any concerns, such as counterintelligence, security, and improper use risks, arise. It also stipulates that an agency may not make operational use of the commercial spyware until at least seven days after providing this information to the White House or until the president's national security adviser consents.

“Ultimately, there will have to be a determination made by the leadership of the department. The outcome may be—based on the information and the facts that we have—that this particular vendor and tool does not spur a violation of the requirements in the executive order,” the senior official says.

While publicly available details of ICE’s contract with Paragon are relatively sparse, its existence alone raised alarms among civil liberties groups, with the nonprofit watchdog Human Rights Watch saying in a statement that “giving ICE access to spyware risks exacerbating” the department’s problematic practices. HRW also questioned what it calls the Biden administration’s “piecemeal approach” to spyware regulation.

The level of seriousness with which the US government approaches the compliance review of the Paragon contract will influence international trust in the executive order, experts say.

“We know the dangers mercenary spyware poses when sold to dictatorships, but there is also plenty of evidence of harms in democracies,” says John Scott-Railton, a senior researcher at the University of Toronto’s Citizen Lab who has been instrumental in exposing spyware-related abuses. “This is why oversight, transparency, and accountability around any US agency attempt to acquire these tools is essential.”

International efforts to rein in commercial spyware are gathering pace. On October 11, during the 57th session of the Human Rights Council, United Nation member states reached a consensus to adopt language acknowledging the threat that the misuse of commercial spyware poses to democratic values, as well as the protection of human rights and fundamental freedoms. “This is an important norm setting, especially for countries who claim to be democracies,” says Natalia Krapiva, senior tech-legal counsel at international nonprofit Access Now.

Although the US is leading global efforts to combat spyware through its executive order, trade and visa restrictions, and sanctions, the European Union has been more lenient. Only 11 of the 27 EU member states have joined the US-led initiative stipulated in the “Joint Statement on Efforts to Counter the Proliferation and Misuse of Commercial Spyware,” which now counts 21 signatories, including Australia, Canada, Costa Rica, Japan, and South Korea.

“An unregulated market is both a threat to the citizens of those countries, but also to those governments, and I think that increasingly our hope is that there is a recognition [in the EU] of that as well,” the senior US administration official tells WIRED.

The European Commission published on October 16 new guidelines on the export of cyber-surveillance items, including spyware; however, it has yet to respond to the EU Parliament's call to draft a legislative proposal or admonish countries for their misuse of the technology.

While Poland launched an inquiry into the previous government’s spyware use earlier this year, a probe in Spain over the use of spyware against Spanish politicians has so far led to no accusations against those involved, and one in Greece has cleared government agencies of any wrongdoing.

“Europe is in the midst of a mercenary spyware crisis,” says Scott-Railton. “I have looked on with puzzled wonderment as European institutions and governments fail to address this issue at scale, even though there are domestic and export-related international issues.”

With the executive order, the US focuses on its national security and foreign policy interests in the deployment of the technology in accordance with human rights and the rule of law, as well as mitigating counterintelligence risks (e.g. the targeting of US officials). Europe—though it acknowledges the foreign policy dimension—has so far primarily concentrated on human rights considerations rather than counterintelligence and national security threats.

Such a threat became apparent in August, when Google’s Threat Analysis Group (TAG) found that Russian government hackers were using exploits made by spyware companies NSO Group and Intellexa.

Meanwhile, Access Now and Citizen Lab speculated in May that Estonia may have been behind the hacking of exiled Russian journalists, dissidents, and others with NSO Group’s Pegasus spyware.

“In an attempt to protect themselves from Russia, some European countries are using the same tools against the same people that Russia is targeting,” says Access Now’s Krapiva. “By having easier access to this kind of vulnerabilities, because they are then sold on the black market, Russia is able to purchase them in the end.”

“It’s a huge mess,” she adds. “By attempting to protect national security, they actually undermine it in many ways.”

Citizen Lab’s Scott-Railton believes these developments should raise concern among European decisionmakers just as they have for their US counterparts, who emphasized the national security aspect in the executive order.

“What is it going to take for European heads of state to recognize they have a national security threat from this technology?” Scott-Railton says. “Until they recognize the twin human rights and national security threats, the way the US has, they are going to be at a tremendous security disadvantage.”

A Russian-aligned propaganda network notorious for creating deepfake whistleblower videos appears to be behind a coordinated effort to promote wild and baseless claims that Minnesota governor and vice presidential candidate Tim Walz sexually assaulted one of his former students, according to several specialists tracking the disinformation campaign.

Experts believe that the campaign is tied to a network called Storm-1516, which has been linked to, among other things, a previous effort that falsely claimed vice president Kamala Harris perpetrated a hit-and-run in San Francisco in 2011. Storm-1516 has a long history of posting fake whistleblower videos, and often deepfake videos, to push Kremlin talking points to the West.

The propaganda unit’s work has successfully reached the highest levels of the Republican party, with vice presidential candidate JD Vance repeating at least one of their narratives. NBC reported this week that the group has pushed at least 50 false narratives in this manner since last fall, which comes amid a broader Russian government effort to disrupt next month’s election with the aim of helping former president Donald Trump return to the White House.

Numerous figures in MAGA world boosted the Tim Walz assault claims, including Jack Posobiec, the Pizzagate promoter who is now a member of Trump’s campaign team, and Candace Owens, the popular right-wing podcaster. The claims went viral on X last week, when an anonymous account called Black Insurrectionist posted screenshots of emails from a purported victim. Other X users quickly debunked the claims, citing formatting errors in the images that suggested the emails were fake, but days later another conspiracist posted a video on X claiming he had spoken to one of Walz's supposed victims on the phone—without providing any proof. The video racked up millions of hits.

Then, on Wednesday, a video claiming to show a former student of Walz describing abuse by the former football coach spread widely on X. According to a WIRED analysis using several deepfake detector tools, the video was created using AI. The video, shared by a prominent anonymous QAnon-promoting account, garnered over 4.3 million views before it was deleted.

The campaign to attack Walz predates the video; it traces back to John Dougan, a former Florida cop who now lives in Moscow and runs a network of pro-Kremlin websites. Dougan appeared on Zak Paine’s QAnon show RedPill78 on October 5 with an anonymous man named “Rick,” who said he was a foreign exchange student at Mankato West High School in 2004 when Walz was a teacher there. “Rick” then claimed Walz assaulted him. Dougan did not respond to a request for comment.

The claims, however, didn’t go viral until last week and the release of the deepfake video.

Darren Linvill, codirector at Clemson University’s Media Forensics Hub, tells WIRED that he immediately recognized this tactic as part of Russia’s well-established disinformation playbook.

“There is little doubt this is Storm-1516,” says Linvill, whose team uncovered the network last fall.

Linvill says the account that first shared the AI-altered video bears all the hallmarks of previous Storm-1516 campaigns. “It is standard for them to create an X or YouTube account for initial placement of stories,” says Linvill.

The campaign orchestrated by Storm-1516 often begins with the posting of a fake story and video from a whistleblower or citizen journalist, the US mission to the Organization for Security and Cooperation in Europe outlined in July. Disinformation is “amplified by other seemingly unaffiliated online networks,” the US mission stated. The claims then take on a life of their own, shared and reposted by unwitting social media users who likely have no idea of where the videos originated.

The fake stories can also be picked up by other media outlets that cover viral social media stories. In the case of the Walz claims, they ended up on MSN, a news aggregation site owned by Microsoft.

In the past, Storm-1516 has relied on a network of fake news websites run by Dougan to push its narratives. On Saturday, a story that referenced the RedPill78 interview, the Black Insurrectionist posts, and the deepfake video was published on over 100 of Dougan’s websites simultaneously.

This was first discovered by Alex Liberty, a researcher who tracks the activity of Russia’s propaganda networks and who agrees with Linvill’s assertion that the deepfake video bears all the hallmarks of a Storm-1516 campaign.

“We believe that it might be a coordinated campaign in [an] attempt to bring numerous false accusations of the same nature against Tim Walz through different channels and in different formats in order to bring an image of legitimacy to the narrative,” Liberty tells WIRED.

McKenzie Sadeghi, the AI and foreign influence editor at NewsGuard, agrees.

“The false narrative appears to be part of a wider campaign pushed by pro-Kremlin media and QAnon influencers ahead of the November 5, 2024, US elections aimed at portraying Walz, whose political appeal is as an everyman schoolteacher and coach, as a pedophile who had inappropriate relationships with minors,” Sadeghi wrote in an analysis of the deepfake video.

From the very beginning, the allegations against Walz were easily debunked. In his interview on the RedPill78 QAnon show, Dougan’s source claimed he was in the US thanks to the State Department–funded Future Leaders Exchange program, which allows students from countries formerly under the control of the Soviet Union the chance to study in the US for a year.

However, a spokesperson for the US State Department, told NewsGuard that it has no record of any Future Leaders Exchange student from Kazakhstan in Mankato area schools from 2000 through 2020. Mankato Area Public Schools communications director Mel Helling told NewsGuard the allegations were “outlandish.”

The baseless claims were shared by some far-right accounts in the days after the episode was published, but they didn’t really take hold until a week later, when the X account known as Black Insurrectionist posted a clip from Dougan’s RedPill78 episode. The clip was viewed over 800,000 times.

Google search trends data shows a huge spike in people searching for “Tim Walz pedophile” and “Tim Walz abuse” on October 13, the day the Black Insurrectionist account began posting their claims.

The Black Insurrectionist account is anonymous and launched a year ago; its followers include Donald Trump Jr. and former Trump adviser Roger Stone. The account’s bio reads: “I am MAGA.” It rose to prominence weeks before the Walz post, when it claimed to have been in contact with a whistleblower at ABC who said Harris had been provided with the questions ahead of her September debate with former president Donald Trump. Those claims were widely debunked by multiple major fact-checking and media organizations.

Last week, the Black Insurrectionist account shared screenshots of email correspondence the account had with an alleged victim on X. Almost immediately, the evidence was questioned when X users spotted a text cursor in one of the screenshots, suggesting that Black Insurrectionist was editing the document. Others pointed out that the date and time format shown in some of the screenshots was inconsistent with how they are displayed on real emails.

Black Insurrectionist initially defended itself before going silent. The account was deleted on Thursday.

The two dozen posts from Black Insurrectionist laying out their alleged evidence have been viewed over 33 million times, according to X’s own metrics, and have been shared on numerous other platforms, including Truth Social, Instagram, Telegram, and TikTok.

Among those sharing Black Insurrectiont’s claims was Paine, who hosted Dougan on his QAnon show. “I have no reason to doubt the veracity of this story,” Paine wrote on X.

The posts have also caught the attention of the wider MAGA universe in a way that Dougan’s initial claims didn’t. Prominent right-wing figures like Owens and Posobiec both flagged the “allegations” as something worth looking into.

Owens discussed the conspiracy on her top-rated podcast, with the episode racking up over 630,000 views on YouTube since it was posted on Wednesday.

Posobiec wrote on X that there were “lots of allegations going around regarding Tim Walz sexually abusing young student(s).” While he added that he didn’t “know about any of the recent allegations being made,” he did share a link to Dougan’s claims from earlier in the month.

When Harris replaced President Joe Biden as the Democratic nominee for president in July, Russian-aligned propaganda networks struggled to mount effective disinformation campaigns targeting the vice president and her team.

But as Microsoft reported in the summer, those campaigns have started to find their footing. "The shift to focusing on the Harris-Walz campaign reflects a strategic move by Russian actors aimed at exploiting any perceived vulnerabilities in the new candidates," Clint Watts, head of Microsoft’s Threat Analysis Center, wrote in August.

With the 2017 unveiling of the ID Buzz concept, Volkswagen announced that the iconic VW bus—forever a symbol of beachy road trips and 1960’s hippie freedom—was returning to the market as an EV. The hype machine went into overdrive.

Jump to 2024, and the vehicle that has been on roads in Europe for about two years is finally, finally, making its way onto US shores.

The cost (around $60K) is maybe higher than many had anticipated, and the vehicle's range (around 230 miles) is likely lower than many had hoped. Throw in a very long wait from the unveiling to its arrival in the marketplace, and the hype has dwindled. After spending a day behind the wheel of one, however, I can say that the ID Buzz has rekindled some of that excitement I felt way back in 2017.

Home on the Range

A VW Bus can't be mentioned without thoughts of tie-dyed Deadheads behind the wheel. Likewise, an EV can't be unveiled to the public without mention (and scrutiny) of its range. The ID Buzz arrives with EPA range numbers that are sure to harsh some mellows: 234 miles for the rear-wheel-drive variant, and 231 miles for the all-wheel-drive model. That’s on par for an electric van but far short of most family vehicles, which can top 300 miles per charge.

During my drive—which took me through San Francisco, across the Golden Gate to Marin County, on some Bay Area freeways, and along the region’s scenic backroads—the RWD version of the ID Buzz averaged 2.8 miles per killowatt-hour over 54 miles. I should note that automotive journalists tend to push vehicles to their limits, testing acceleration, stopping frequently, judging the handling, and so on. Based on the vehicle's 91 kWh gross capacity pack (of which, 86 kWh is available), the van was posting 240 miles of range. In regular use, it's likely it will hit 245 to 250 miles of range.

That discrepancy between reported range and real-world range isn't completely unexpected, as the Volkswagen Group has a history of underreporting its numbers. It did so with the Audi E-Tron, Porsche Taycan, and VW ID 4.

The AWD version posted an even more impressive 3.0 miles per kWh, bringing its range up to 258 miles, as measured over 40 miles. During this portion of the day's drive, more of the route was on freeways; I experienced moderate traffic during this period and rarely got above 65 miles an hour. In fact, my average speed was 42 miles per hour.

This real-world range should be more than adequate for hauling people and cargo around town. For road trips and camping—what it feels made for—the vehicle's very quick 200-kW peak charging rate (via a 400-volt architecture) does help reduce the pain of a sub-250-mile EPA rating. I'd rather have a vehicle with ID Buzz's battery capacity with 230 miles of range that charges up at a super-fast 200 kW than a vehicle with 300 miles of range that charges at a much slower 120 kW.

For those charging at home with a wall socket, the vehicle supports AC charging at up to 11 kW.

The Transporter

If you've ever had the pleasure of piloting an old-school VW bus (aka Samba, Bulli, Transporter, or Microbus), then you're aware how painfully slow it is to accelerate. It feels like the van’s zero-to-60 time is about three days. The ID Buzz's electric drivetrain solves that issue, even if it's never going to impress your Hyundai Ioniq 5N–owning friends. The “speedy” AWD trim level with 335 horsepower and 512 pound-feet of torque can motor from zero to 60 in about six seconds.

In a world where every EV needs to be able to pull a quarter mile in under 13 seconds, the leisurely acceleration of the ID Buzz is not only appropriate but welcome. It's a van. It hauls your whole family. Undue acceleration makes passengers uncomfortable and scatters your Kirkland Signature sundries around in the cargo area.

I found the added acceleration helpful in a few instances while overtaking other drivers, but the traction (combined with appropriate all-season or winter tires) will make the AWD ID Buzz a solid winter vehicle for those living in regions with snow and ice.

For a daily driver, I prefer the lighter RWD variant with its 282 horsepower and 413 pound-feet of torque. It's more engaging around corners, and the reduced heft of the vehicle gives it a more nimble feel. To be clear, the Buzz is not a canyon-carving machine, but I was impressed by its ability to handle the curves along the Bay Area backroads. As with all EVs, the low center of gravity afforded by the battery pack helps. For its intended purpose as a people mover though, the vehicle's suspension feels just a bit too stiff. Sure, that helps with cornering, but it translates to a slighter bumpier ride.

Slow Down There, Bulli

One interesting aspect of the ID Buzz is how Volkswagen handles passive regenerative braking. In the default comfort mode, it does not use passive regen. You can lift your foot off the accelerator and just coast. If you turn the steering column-mounted shifter to B, passive regen is engaged. It feels about mid-level compared to the similar feature on other vehicles; the van slows but not too much. There is no option for the type of one-pedal driving found in EVs from BMW, Chevrolet, and Tesla.

You can also access an even less aggressive passive regen when the vehicle is placed into Sport mode. Volkswagen says it's there to re-create the feeling of engine braking. The problem here is that all the driving modes are located in the vehicle settings portion of the infotainment system. Changing the drive mode requires two taps on the main screen. Sport mode makes only a slight adjustment to the acceleration mapping to increase the Buzz’s acceleration speed, so it's not really worth the hassle. Unless you turn this on before you embark, stick with Comfort mode.

So Much Room for Activities

The hype around the ID Buzz proves the power of nostalgia. And since the automotive world loves celebrating the past, automakers exploit that feeling in their designs. The DNA of those original VW microbuses is apparent in the ID Buzz. It has short overhangs, tons of windows, and a two-tone paint job. It's the microbus we never bought (but maybe wanted to) wrapped in a greener package.

Thankfully, just like the original Microbus, the ID Buzz is a wonder of utility. It has a whopping 146 cubic feet of cargo space. That's one cubic foot more than a Chevy Suburban and five more than the Chrysler Pacifica plug-in hybrid. The RWD version comes standard with seating for seven. The AWD variant ships with captains chairs in the second row and seats six. All three rows can be slid forward or back to adjust for legroom and storage.

Unlike in most three-row SUVs, sitting at the back of the Buzz was a comfortable experience for my 6-foot-3 body. We also piled four WIRED colleagues into the van for a long loop around the neighborhood, and everyone had plenty of room.

With all the seats in use, the cargo space behind the third row is 18.6 cubic feet. That’s enough for a large grocery haul or several peoples’ luggage. The third row of seats can be folded down to make a flat surface in the back, and VW offers an insert for the rear cargo area that comes with two handy drawers. If you need more space, the third row can be removed entirely.

To help keep items from sliding around, the Buzz is equipped with velcro partitions that are stored in the walls of the cargo area. The center console is removable and moveable. You can move it from between the front seats to the second row in vehicles equipped with captain's chairs. Or just pull it out entirely. One nice touch: The dividers in the center console double as a bottle opener and an ice scraper.

Meanwhile, VW has gone full USB-C. Every seat gets at least one charging port. There are seven ports in total. There’s also one 15-watt USB port near the rearview mirror to accommodate a dash cam. A 110-volt, 150-watt power outlet is available under the passenger seat.

This all before you get into the world of aftermarket add-ons, where you are sure to find a plethora of options for the Buzz. It's all very clever, and a reminder that the VW bus is a canvas for your lifestyle.

Finally, the issues that have famously been plaguing VW’s infotainment system for years have been solved. The 12.9-inch display was easy to use, navigate, and more importantly, had little in the way of latency. VW has added ChatGPT integration for the voice assistant, but that requires a network connection, and I wasn’t able to fully test it, since most of the drive was in areas where cell service was spotty or absent. When I was able to test it, it returned a solid answer in a timely fashion.

Weirdly though, as an adventure vehicle, the Buzz does not ship with a dog or camping mode. When asked about this, Volkswagen said it was looking into it. More than a few journalists inquired about these features, which are found in Rivians and Teslas and make all sorts of sense for the Buzz. So don’t be surprised if those modes show up in an OTA software update.

How Much Again?

Which brings us to the second sticking point of the ID Buzz. The RWD Pro S model starts at $59,995, and this is because Volkswagen essentially offers a mid-level trim as the entry-level model. For example, this starting configuration ships with 12-way adjustable heated and vented front seats with a massage feature. Heated seats are also standard for the second row. It has three-zone climate controls.

All those USB ports are also standard. So is that 110V outlet. For those late night raves, a 30-color ambient lighting feature is standard.

Volkswagen is building all its ID Buzzes at a single factory in Hanover, Germany. This means US buyers cannot claim an EV purchase tax credit, since the Inflation Reduction Act requires vehicles to be assembled in North America to qualify for the tax break. If the automaker offers a lease, then the van does become eligible.

Volkswagen is aware this is a niche vehicle. While the buzz around the Buzz has died down, it will still likely sell out in its first year. VW would not comment on whether it will offer a cheaper trim option in the future or whether it plans to expand its production to the US.

Yet even with its lack of a true entry-level price and a range that, while underreported, is less than anticipated, the ID Buzz is exactly what it should be. It is a fun-to-drive nostalgia machine with enough storage and utility to make it a solid weekend hauler for families who enjoy an active lifestyle. And your five children will be quite comfortable in the back as you bore them to tears with your Summer of Love playlist.

Kwikset is no newbie in the lock game. The company has been around since the 1940s making your classic locks and doorknobs and has been making smart locks under its Halo line since 2019. The Halo line originally consisted of two locks, the Halo Keypad and the Halo Touch. As of today there’s a third: the Halo Select.

The Halo Select packs a variety of updates and changes from the previous Halo models, including a quieter motor, unlocking via geofence, and a brand-new door sensor. It's also the brand's first lock compatible with Matter, the smart-home standard that allows smart-home devices to communicate without requiring multiple apps and hubs.

It took Kwikset a while to come out with a Matter lock after companies like Yale did it first. “This was not an easy feat for us," said Charlie Doughtery, Kwikset's electronics brand manager. "We as a company wanted to put out something that was great for consumers and offered a best-in-class consumer experience, and really made sure we got there before we decided to release to market.”

A lot of these features we've already seen in smart locks. But based on what we've seen so far, the Matter experience is one of the best I've seen, and the super-minimal door-status sensor is a particularly refreshing take in the world of smart sensors. More on that below.

Matter Time

This isn't the first Matter smart lock, but the Matter experience that Kwikset designed into its app is pretty impressive.

Now, the whole point of Matter is that you shouldn't need separate apps to manage all of your devices. Kwikset still requires you to start with its app for the initial setup, but then you'll head into the Lock Settings and tap on the section labeled Matter, where you can easily switch over control to your smart-home ecosystem of choice. The app will prompt you to choose Amazon Alexa, Google Assistant, Apple Home, or Samsung SmartThings when you switch over to Matter, and then will hand over the keys to your choice of ecosystem.

After testing a few Matter devices, I don't mind starting with a brand's app, and I think it especially makes sense when installing a smart lock as opposed to, say, setting up a smart plug or smart light bulb. There are several more steps involved when you install a smart lock, and getting it wrong can mean no working lock whatsoever on your door.

The Kwikset app will also make it easy to switch back control to Kwikset and Wi-Fi if you decide Matter isn't for you. There are more features available if you choose the Kwikset-controlled path, like guest codes and using the included door-status sensor, but the lock does promise better battery life when using Matter.

Slim Sensing

The most exciting addition to Kwikset's newest lock is the included door-status sensor. It works like any other door sensor—the sensors line up to tell you whether the door is closed—but it's a super-slim, discreet style compared to other options on the market.

Kwikset hasn't had door sensors before, but other smart-lock brands like Yale (which now owns August locks) have offered these for a while in conjunction with their locks, and plenty of other smart-home gadget makers have offered stand-alone door and window sensors for a similar experience. In both cases, it's a two-part sensor that sits on your doorframe or window frame and door or window, and usually requires a few sticky cushions to perfectly line the two parts up so that the sensor can correctly identify when the door is opened or closed.

Kwikset won't use that system. Instead, the brand designed a super-slim magnet-style sensor that you'll place into the doorjamb where it can connect with the lock. It looks incredibly sleek and seems like it would be much easier to place than trying to perfectly line up two separate sensors on a door and frame. Only testing will tell if this is indeed the case, and it's top of my testing list once it becomes available.

The Kwikset Halo doesn't have a specific launch date but is expected to be available at retailers like Lowe's and Amazon by mid-November and will retail for $279.

On a dead-end road that climbs out of the tiny city of Jenkins, in the foothills of the Appalachian Mountains in Eastern Kentucky, there stands a large warehouse with a mint green roof. It shares the road with a few other businesses, but is otherwise surrounded by an expanse of open fields and tree-lined slopes. Inside, the warehouse is stacked high with racks on racks of computers—thousands of them. But none have ever been switched on.

The warehouse is owned by Mohawk Energy, a company cofounded by Kentucky state senator Brandon Smith in 2005, originally to resculpt landscapes disfigured by coal mining. After lying dormant for a period, Mohawk was reincarnated in 2022 when Smith struck a deal with HBTPower, a company then owned by Chinese crypto exchange Huobi, which wanted to use the warehouse for a bitcoin mining operation.

Under the deal, Mohawk promised to fit up its warehouse with the necessary power infrastructure, operate the equipment, and funnel any bitcoin produced to HBT. In return, HBT would pay Mohawk a monthly hosting fee, a cut of its mining revenue, and the associated energy bills.

Smith says he hoped the arrangement would generate tax revenue and create jobs for former coal miners, who could be trained as repair technicians. The coal industry departed Jenkins long ago, the reserves depleted, leaving people in search of work. More than a third now live below the poverty line, per the latest census data. “I liked the idea of going from one type of mining to a new type,” says Smith. “I thought, now in Eastern Kentucky we are going to have our time—we’re going to catch up and play a part in the tech future.”

But after a promising start, the relationship between Mohawk and HBT soured and then fell apart. “Nothing has ever been turned on. It’s a fascinating, almost Willy Wonka–type atmosphere when you walk through,” says Smith. “It has turned into a disaster.”

In November 2023, HBT brought a lawsuit in federal court, alleging that Mohawk had breached its contract on several fronts, including by failing to install the appropriate power infrastructure and secure certain power subsidies, and attempting to sell off the mining equipment. “Ultimately, the source of the current dispute is Mohawk’s basic failure to comply with its obligations, not only in a timely way, but at all in many regards,” says Harout Samra, a specialist in international dispute resolution at law firm DLA Piper and representative for HBT.

Mohawk sued HBT in return, contesting the various alleged breaches and claiming that HBT is delinquent on more than $700,000 in rent, labor, and fit-up costs. The company is also seeking damages relating to the loss of income over the term of the contract and the inability to bring a new tenant into the facility while the equipment remains on-site. “Huobi simply made a bargain it believes now is a bad one, and wants to get out of it without paying the funds it owes,” the filing states.

The legal conflict, which remains unresolved, is just one in a series of fights between Chinese companies and the owners of industrial facilities in the rural US over failed bitcoin mining partnerships. What looked to facility owners in Kentucky like an irresistible opportunity to tap into a new line of business in an otherwise fallow period has turned into a nightmare. They claim to have been saddled with unpaid hosting fees and energy bills worth hundreds of thousands of dollars, with few options for recovering the money. The Chinese parties have been left equally displeased. “HBTPower obviously regrets that this opportunity has ultimately played out the way it has,” says Samra.

The bitcoin mining game—a race between computers to win the right to process a bundle of transactions and claim a crypto reward—is dominated by large corporations that own and operate industrial-scale facilities. But in 2021 and 2022, smaller-scale operations began to proliferate in the US countryside wherever there was available power, including in Kentucky. “A lot of mom-and-pop shops opened up,” says Phil Harvey, CEO at Sabre56, a firm that consults on crypto mining projects and operates its own facilities. “Appalachia has always been a good source of power.”

These small facilities were plugging a gap in the market. A ban on crypto mining in China had left businesses casting about for a new home for their many millions of dollars’ worth of mining equipment. “A lot of wealthy Chinese businesses were affected,” says Harvey. “Every minute these machines are down, they are losing revenue.” Meanwhile, as the price of bitcoin ballooned—and the profitability of mining along with it—mining firms and investor groups began to hoard large quantities of bitcoin mining equipment of their own, says Harvey, without considering where they might deploy it.

In an overheated market, holders of mining equipment jumped into hosting arrangements at short notice with owners of small facilities, some of whom had no prior experience and insufficient expertise, who agreed to install the equipment and run the mining operations on their behalf.

But the haste with which these hosting relationships came together, in the name of striking while bitcoin was hot, says Harvey, set many of the partnerships up for failure. There was limited due diligence conducted by parties on both sides, delays in kitting out facilities and deploying equipment, and disputes over payment terms, he says, among other points of friction. “It's a snowball effect where everyone just ends up getting pissed off with each other,” says Harvey.

Though the American market proved more expensive and bureaucratic than some Chinese businesses expected, says Harvey, problems were also caused by the hubris of facility owners, some of whom found themselves in over their heads. “It’s no joke running a [bitcoin mining] operation of any kind of scale,” he says. “Just because the Chinese are tough to do business with, doesn’t mean they are the ones in the wrong. I would say that blame is equally shared.”

The law firm acting for Mohawk in its dispute with HBT, Anna Whites Law Office, has represented multiple owners of small facilities in Kentucky in similar legal conflicts with Chinese partners. The cases differ from the Mohawk situation, says attorney Anna Whites, founder of the firm, but share a common thread: “We saw a pattern that [companies with ties to China] would ship in machines with uncertain provenance, mine very heavily for three months, then run without paying the bill,” she claims.

Some of the cases settled out of court; Whites is unable to supply the details for reasons of client confidentiality. But others continue to drag on.

Biofuel Mining, a company formerly co-owned by Smith, is involved in legal tangles with two companies that Whites believes to be run out of China: Touzi Tech and VCV Power Gamma. Although both are incorporated in Delaware, per SEC filings, they conduct business in Mandarin and cannot be reached at their listed US addresses, Whites claims. “It's pretty standard for the foreign entities from any country to get a short-term office so that they have less scrutiny from US investors and government agencies,” she says.

In both cases, Biofuel claims, the firms shipped equipment from China to its hosting facility in Eastern Kentucky, then walked away with the bitcoin produced, leaving behind hundreds of thousands of dollars in unpaid energy bills and hosting fees.

Biofuel reached a settlement with Touzi in early 2022 for $60,000, but despite having handed back the mining equipment, it claims not to have received the sum it is owed under the agreement.

In the still-unresolved spat with VCV, Biofuel received permission from the Martin County Circuit Court in Kentucky to sell off the mining equipment, claims Whites, to recoup a portion of the funds it is owed (she has not confirmed the amount), but she alleges that no damages have yet been awarded. VCV has stopped responding to communications, she claims.

Biofuel has since dissolved, put out of business by the failed hosting ventures. “I literally lost my house—I lost everything. It financially ruined me,” says Wes Hamilton, former Biofuel Mining CEO. “I’m just so frustrated about the whole thing.”

WIRED contacted VCV and Touzi for comment, but did not receive any response.

There are few financial recovery options for companies like Mohawk and Biofuel. The situation is made more difficult, as in the Mohawk case, if they are dealing with so-called special purpose entities. Because they are set up by their parent companies for a single specific business venture, these entities need not be concerned about their long-term ability to operate in the US.

“It certainly can be more difficult to recover damages from a non-US counterparty,” says Kim Havlin, a partner in the global commercial litigation practice at law firm White & Case. “There is certainly a risk that an entity that doesn’t need to be in the US may just ignore the case.”

Even if the Kentucky facility owners win out in court, it could be difficult to collect any damages awarded. “A judgment is essentially a piece of paper. Any judgment needs to be turned into assets or cash in order to be valuable,” says Havlin. If the opposing party refuses to pay up and has no US assets to collect against, sometimes that isn’t possible.

Almost a year after the dispute began, the Mohawk case is stuck in legal limbo. In a setback for Mohawk, the presiding judge recently denied its motion to dismiss HBT’s complaint, on the basis that it had failed to sufficiently back up its argument. The judge also pushed Mohawk’s countersuit into arbitration, a forum for resolving disputes privately instead of in open court. Non-US parties tend to prefer arbitration as a way to “remove a home forum from both sides,” explains Havlin. “You can pick an arbitral seat in neither country as a means of creating a neutral playing field.” A parallel federal court hearing is set for December to consider whether an injunction should be imposed on Mohawk, preventing it from selling off the remaining HBT equipment in its possession.

Smith has given up on the idea of recovering the full amount he claims to be owed. “We’re at the point that it’s almost silly to even be arguing about breaking even,” he says.

In an interview with PBS that aired in September 2023, touting the Mohawk Energy facility, Smith said he hoped to prove that not every business that blew into Jenkins would abandon the area. “I’ve stood at their ribbon cuttings, then watched them leave. I’d like to do something to let people know that not everybody is like that,” he said.

After the relationship with HBT collapsed last year, Smith faces the prospect of Mohawk becoming yet another false start. With the facility inactive, the company has been forced to dismiss the former coal miners brought on as technicians. (It is unclear how many people it employed.)

The Mohawk facility was perhaps never set to revitalize Jenkins in the way Smith hoped, anyway. “I would say that a rural community benefits very little from a bitcoin mining facility. In terms of job creation, it’s minimal in a lot of cases,” says Harvey, the consultant. “It's certainly not the savior to a dwindling community.”

Nonetheless, Smith remains hopeful of salvaging the crypto mining project, with a new partner. “I’m hoping that this gets settled in the way that it should and that somebody comes forward and lets us go through with the vision that we wanted for this region,” he says. “I hope every day that maybe some big company will see that there's a place ready to go in this part of the country.”

Otherwise, Mohawk’s dalliance with bitcoin mining will become a cautionary tale. “It was very hurtful to see these families lose their income. We were one of the biggest payrolls in Jenkins,” says Smith. “It adds insult to injury that I’m sitting here arguing in court.”

If Donald Trump wins the US presidential election in November, the guardrails could come off of artificial intelligence development, even as the dangers of defective AI models grow increasingly serious.

Trump’s election to a second term would dramatically reshape—and possibly cripple—efforts to protect Americans from the many dangers of poorly designed artificial intelligence, including misinformation, discrimination, and the poisoning of algorithms used in technology like autonomous vehicles.

The federal government has begun overseeing and advising AI companies under an executive order that President Joe Biden issued in October 2023. But Trump has vowed to repeal that order, with the Republican Party platform saying it “hinders AI innovation” and “imposes Radical Leftwing ideas” on AI development.

Trump’s promise has thrilled critics of the executive order who see it as illegal, dangerous, and an impediment to America’s digital arms race with China. Those critics include many of Trump’s closest allies, from X CEO Elon Musk and venture capitalist Marc Andreessen to Republican members of Congress and nearly two dozen GOP state attorneys general. Trump’s running mate, Ohio senator JD Vance, is staunchly opposed to AI regulation.

“Republicans don't want to rush to overregulate this industry,” says Jacob Helberg, a tech executive and AI enthusiast who has been dubbed “Silicon Valley’s Trump whisperer.”

But tech and cyber experts warn that eliminating the EO’s safety and security provisions would undermine the trustworthiness of AI models that are increasingly creeping into all aspects of American life, from transportation and medicine to employment and surveillance.

The upcoming presidential election, in other words, could help determine whether AI becomes an unparalleled tool of productivity or an uncontrollable agent of chaos.

Oversight and Advice, Hand in Hand

Biden’s order addresses everything from using AI to improve veterans’ health care to setting safeguards for AI’s use in drug discovery. But most of the political controversy over the EO stems from two provisions in the section dealing with digital security risks and real-world safety impacts.

One provision requires owners of powerful AI models to report to the government about how they’re training the models and protecting them from tampering and theft, including by providing the results of “red-team tests” designed to find vulnerabilities in AI systems by simulating attacks. The other provision directs the Commerce Department’s National Institute of Standards and Technology (NIST) to produce guidance that helps companies develop AI models that are safe from cyberattacks and free of biases.

Work on these projects is well underway. The government has proposed quarterly reporting requirements for AI developers, and NIST has released AI guidance documents on risk management, secure software development, synthetic content watermarking, and preventing model abuse, in addition to launching multiple initiatives to promote model testing.

Supporters of these efforts say they’re essential to maintaining basic government oversight of the rapidly expanding AI industry and nudging developers toward better security. But to conservative critics, the reporting requirement is illegal government overreach that will crush AI innovation and expose developers’ trade secrets, while the NIST guidance is a liberal ploy to infect AI with far-left notions about disinformation and bias that amount to censorship of conservative speech.

At a rally in Cedar Rapids, Iowa, last December, Trump took aim at Biden’s EO after alleging without evidence that the Biden administration had already used AI for nefarious purposes.

“When I’m reelected,” he said, “I will cancel Biden’s artificial intelligence executive order and ban the use of AI to censor the speech of American citizens on Day One.”

Due Diligence or Undue Burden?

Biden’s effort to collect information about how companies are developing, testing, and protecting their AI models sparked an uproar on Capitol Hill almost as soon as it debuted.

Congressional Republicans seized on the fact that Biden justified the new requirement by invoking the 1950 Defense Production Act, a wartime measure that lets the government direct private-sector activities to ensure a reliable supply of goods and services. GOP lawmakers called Biden’s move inappropriate, illegal, and unnecessary.

Conservatives have also blasted the reporting requirement as a burden on the private sector. The provision “could scare away would-be innovators and impede more ChatGPT-type breakthroughs,” Representative Nancy Mace said during a March hearing she chaired on “White House overreach on AI.”

Helberg says a burdensome requirement would benefit established companies and hurt startups. He also says Silicon Valley critics fear the requirements “are a stepping stone” to a licensing regime in which developers must receive government permission to test models.

Steve DelBianco, the CEO of the conservative tech group NetChoice, says the requirement to report red-team test results amounts to de facto censorship, given that the government will be looking for problems like bias and disinformation. “I am completely worried about a left-of-center administration … whose red-teaming tests will cause AI to constrain what it generates for fear of triggering these concerns,” he says.

Conservatives argue that any regulation that stifles AI innovation will cost the US dearly in the technology competition with China.

“They are so aggressive, and they have made dominating AI a core North Star of their strategy for how to fight and win wars,” Helberg says. “The gap between our capabilities and the Chinese keeps shrinking with every passing year.”

“Woke” Safety Standards

By including social harms in its AI security guidelines, NIST has outraged conservatives and set off another front in the culture war over content moderation and free speech.

Republicans decry the NIST guidance as a form of backdoor government censorship. Senator Ted Cruz recently slammed what he called NIST’s “woke AI ‘safety’ standards” for being part of a Biden administration “plan to control speech” based on “amorphous” social harms. NetChoice has warned NIST that it is exceeding its authority with quasi-regulatory guidelines that upset “the appropriate balance between transparency and free speech.”

Many conservatives flatly dismiss the idea that AI can perpetuate social harms and should be designed not to do so.

“This is a solution in search of a problem that really doesn't exist,” Helberg says. “There really hasn’t been massive evidence of issues in AI discrimination.”

Studies and investigations have repeatedly shown that AI models contain biases that perpetuate discrimination, including in hiring, policing, and health care. Research suggests that people who encounter these biases may unconsciously adopt them.

Conservatives worry more about AI companies’ overcorrections to this problem than about the problem itself. “There is a direct inverse correlation between the degree of wokeness in an AI and the AI's usefulness,” Helberg says, citing an early issue with Google’s generative AI platform.

Republicans want NIST to focus on AI’s physical safety risks, including its ability to help terrorists build bioweapons (something Biden’s EO does address). If Trump wins, his appointees will likely deemphasize government research on AI’s social harms. Helberg complains that the “enormous amount” of research on AI bias has dwarfed studies of “greater threats related to terrorism and biowarfare.”

Defending a “Light-Touch Approach”

AI experts and lawmakers offer robust defenses of Biden’s AI safety agenda.

These projects “enable the United States to remain on the cutting edge” of AI development “while protecting Americans from potential harms,” says Representative Ted Lieu, the Democratic cochair of the House’s AI task force.

The reporting requirements are essential for alerting the government to potentially dangerous new capabilities in increasingly powerful AI models, says a US government official who works on AI issues. The official, who requested anonymity to speak freely, points to OpenAI’s admission about its latest model’s “inconsistent refusal of requests to synthesize nerve agents.”

The official says the reporting requirement isn’t overly burdensome. They argue that, unlike AI regulations in the European Union and China, Biden’s EO reflects “a very broad, light-touch approach that continues to foster innovation.”

Nick Reese, who served as the Department of Homeland Security’s first director of emerging technology from 2019 to 2023, rejects conservative claims that the reporting requirement will jeopardize companies’ intellectual property. And he says it could actually benefit startups by encouraging them to develop “more computationally efficient,” less data-heavy AI models that fall under the reporting threshold.

AI’s power makes government oversight imperative, says Ami Fields-Meyer, who helped draft Biden’s EO as a White House tech official.

“We’re talking about companies that say they’re building the most powerful systems in the history of the world,” Fields-Meyer says. “The government’s first obligation is to protect people. ‘Trust me, we’ve got this’ is not an especially compelling argument.”

Experts praise NIST’s security guidance as a vital resource for building protections into new technology. They note that flawed AI models can produce serious social harms, including rental and lending discrimination and improper loss of government benefits.

Trump’s own first-term AI order required federal AI systems to respect civil rights, something that will require research into social harms.

The AI industry has largely welcomed Biden’s safety agenda. “What we're hearing is that it’s broadly useful to have this stuff spelled out,” the US official says. For new companies with small teams, “it expands the capacity of their folks to address these concerns.”

Rolling back Biden’s EO would send an alarming signal that “the US government is going to take a hands off approach to AI safety,” says Michael Daniel, a former presidential cyber adviser who now leads the Cyber Threat Alliance, an information sharing nonprofit.

As for competition with China, the EO’s defenders say safety rules will actually help America prevail by ensuring that US AI models work better than their Chinese rivals and are protected from Beijing’s economic espionage.

Two Very Different Paths

If Trump wins the White House next month, expect a sea change in how the government approaches AI safety.

Republicans want to prevent AI harms by applying “existing tort and statutory laws” as opposed to enacting broad new restrictions on the technology, Helberg says, and they favor “much greater focus on maximizing the opportunity afforded by AI, rather than overly focusing on risk mitigation.” That would likely spell doom for the reporting requirement and possibly some of the NIST guidance.

The reporting requirement could also face legal challenges now that the Supreme Court has weakened the deference that courts used to give agencies in evaluating their regulations.

And GOP pushback could even jeopardize NIST’s voluntary AI testing partnerships with leading companies. “What happens to those commitments in a new administration?” the US official asks.

This polarization around AI has frustrated technologists who worry that Trump will undermine the quest for safer models.

“Alongside the promises of AI are perils,” says Nicol Turner Lee, the director of the Brookings Institution’s Center for Technology Innovation, “and it is vital that the next president continue to ensure the safety and security of these systems.”

Analysts at the US Department of Homeland Security shared an internal report to local agencies in August, warning them about the economic risks of using Chinese utility storage batteries. It warns that the dependence on Chinese batteries could hurt developing a secure supply chain in the US.

The document, first obtained by national security transparency nonprofit Property of the People and seen by WIRED, accuses Chinese companies of “using People’s Republic of China state support to quickly and cheaply enter the emerging US utility battery energy storage industry and create supply chain dependencies on China,” and asks that any suspicious activity be reported.

Specifically, the report alleges three companies—Contemporary Amperex Technology Co. Limited (CATL), Build Your Dreams (BYD), and Ruipu Energy Co. Ltd. (REPT)—have “benefited from the various forms of state support and leveraged this to further business strategies for gaining US market share.”

Currently, CATL and BYD lead the global energy storage battery market by far, with 40 percent and 12 percent market shares, respectively, according to South Korean energy research firm SNE Research. Eight out of the 10 top companies in the industry are from China, so there are few alternatives to turn to when building grid storage.

The report says it builds on previous documents that analyzed Chinese “state-supported firms’ use of noncompetitive tactics in the electric vehicle and battery supply chains.” DHS did not respond to a request for further comment.

In 2022, CATL entered a deal with Primergy Solar to build the largest US solar and storage project in Nevada, which came online this year. Its battery products have also been used by Duke Energy, a North Carolina–based utility company, although the latter dropped CATL as a supplier for marine base electricity storage after concerns around national security were raised by, in part, lawmakers in Washington.

In an emailed statement, Fred Zhang, a CATL spokesperson, rejects the categorization that the firm has relied on state support to gain an edge. “CATL has achieved tremendous growth through continuous innovation, farsighted strategic planning, and a commitment to high-quality products at a reasonable cost,” the statement says.

BYD and REPT did not reply to WIRED’s request for comment.

Following efforts to curb Chinese EV companies’ competitiveness, the US government is now also concerned about how domestic utility companies could become too dependent on Chinese batteries for energy storage.

The US government has in recent years started to catch up in the battery industry. The Inflation Reduction Act and the Bipartisan Infrastructure Bill set out investment tax credits and other economic incentives to build up energy storage capacity in the country. In September it awarded another $3 billion in incentives to projects that boost domestic production of batteries.

These incentives are also the focus of the DHS report, which alleges that Chinese-based firms are targeting US incentives to further grow their market share. “BYD’s public website as of spring 2023 highlighted how US state incentives can make BYD utility storage systems an attractive investment,” the report claims. (WIRED was unable to find the BYD website referenced in the report.)

The CATL spokesperson says the company “does not receive any US federal or state incentives.”

As the US utility grids incorporate more renewable energy sources like solar and wind, it’s essential to build up a battery storage capacity that can store intermittent energy supply for times of heightened demand. And Chinese companies have dominated the global industry of producing lithium batteries for this job. These companies make over 80 percent of the EV battery cells in the world, and they had plans to invest another $2 trillion in 2023 into new production capacities inside and outside the country.

“Chinese original equipment manufacturers currently supply about 90 percent of the energy storage system batteries, actually a larger share than for EVs,” says Vanessa Witte, a senior research analyst at Wood Mackenzie who focuses on US energy storage. “There is a huge oversupply right now, partly due to softening EV demand, along with a number of Tier 2 and Tier 3 OEMs that have started new manufacturing facilities. The excess supply is a big reason the prices are so low.”

These batteries are essential for global energy transition, and Chinese battery companies see them as an opportunity to widen their product markets. They’ve become such an important industry that the Chinese central government emphasized their importance for the first time in its annual government report in March.

The battery supply chain has also emerged as one of the top economic and security concerns around China in the eyes of the US government. The Biden administration has so far raised a 25 percent tariff on Chinese-made lithium-ion EV batteries (and even higher for the EVs they power.) The government has also repeatedly sounded alarms about the national security threats of having Chinese-made equipment and parts in domestic infrastructure.

So far, the particular conversations around energy storage batteries have mostly surrounded cybersecurity, worried that the components could contain backdoor access for hacking like those that have been suspected in Chinese-made port cranes. Those concerns are “a bit overstated,” Witte says. “There haven’t been any incidents that would lend one to believe the battery management system is sending data to China or could disrupt our infrastructure.”

There have been several political efforts to restrict the use of Chinese-made batteries in the US. The pentagon will be the first to be banned from buying batteries from six Chinese companies, in an order which will become effective in 2027, but some congressmembers are still asking for CATL and other Chinese companies to be added to trade blacklists, and there’s a bill in Congress that would ban DHS from procuring Chinese batteries.

But the DHS report shows that the government is also looking at whether the dominance of Chinese batteries can be an economic issue.

“It is simplistic to see this as just unfair competition due to Chinese government support, as a lot of what China did, such as creating demand for batteries through EV subsidies, is not unlike what we see in the West today,” says Yayoi Sekine, head of energy storage research at BloombergNEF.

The pricing advantage that Chinese battery companies have are also the results of harsh market competition and a more established battery supply chain in China, she says, and these battery companies are willing to squeeze their own margins in exchange for keeping their market shares. “We think the US government's attempts to support a healthy battery industry domestically has been incredibly generous through the incentives in the Inflation Reduction Act and the Bipartisan Infrastructure Law,” says Sekine, “but it may still be challenging to compete on cost.”

It tends to really be used only by developers, but the address bar and search box up at the top of the Google Chrome interface has an official name: the omnibox. It reflects the multipurpose capabilities of this little text field, as it's able to do much more than look up web addresses and run searches on Google.

When you know about everything the omnibox can do, you can save time jumping between different apps and sites, and get things done more quickly. What's more, Google is constantly adding new features to the omnibox. Most recently, as you might expect, the company added an integration with Gemini AI.

Here are a few of our favorites—just remove the quotes around the text examples below to get the code you need to type into the omnibox.

Chat With Gemini

We've just mentioned the most recent upgrade to the Chrome omnibox, so we may as well start here: Type out "@gemini" in the address bar, then a space, then your prompt for the chatbot. Hit Enter, and the query will be run in Google Gemini. Chrome will use whatever flavor of Gemini is included with your Google account (so Gemini Advanced, if you're a paying user).

Carry Out Conversions

Any kind of conversion you need, the all-powerful omnibox can take care of for you: Turn kilometers into miles, or dollars into euros, or days into months. All you have to do is type out the desired conversion in a way that makes sense. Chrome is pretty good at working out what you're trying to do, so for example, you can type “£34 in us dollars” and it will know you’re looking for a conversion. You should immediately see the result appear underneath—you don't need to hit Enter.

Run Basic Calculations

On a related note, you can run simple calculations from the Chrome omnibox as well, no need to press Enter. Anything like "24*8" or "352+91" will instantly show a result underneath—as will "24*8-352+91"—and you can use brackets if you need part of the sum worked out first. If you do press Enter afterwards, the full Chrome calculator opens up.

Check the Weather

Chrome can report on live weather conditions from the omnibox. Just type "weather" (no need to press Enter) to see a mini description of the current conditions in wherever you are. Note that this only gives the most accurate result if Chrome has access to your current location. Add a town, city, or postal code on the end to see conditions in that place, and hit Enter after your query for a more detailed forecast.

Search Your Bookmarks

You can search through your Chrome bookmarks right from the omnibox, without having to open up the browser’s integrated Bookmark Manager. You do need to type out the name of one of your bookmark folders first, so Chrome knows what you're trying to do, and you can then write any word or phrase to see instant results for pages saved in that bookmarks folder.

Make Notes in Chrome

If you need to quickly get some thoughts down in Chrome and you don't want to launch a separate program, the code "data:text/html, <html contenteditable>" followed by Enter will give you a blank tab you can type into. It's not the most advanced of text editors—there's no formatting and no auto-save—but it works well as a quick solution for jotting down notes.

Get Quick Definitions

If you're unsure what a particular word means, Chrome can tell you, and you don't need to leave the page you're currently on to find out the definition. Type "define", then a space, then the word you want the meaning for, and a basic definition pops up underneath. To get back to the URL of the page you were viewing, press Esc to remove the definition.

Create New Documents

You can quickly create new documents, spreadsheets, or presentations in Google's online office suite by typing "docs.new", "sheets.new", or "slides.new" into the omnibox. When you press Enter, the new file is created in the Google Drive for the current Google account. To create a new file in a new window (leaving the current one alone), use Shift+Enter after your command.

There's a whole suite of things .new shortcuts can launch, and Google continues to add new capabilities.

Start New Emails

There's a similar trick for creating new emails in the default email client on your computer: Type "mailto:" and hit Enter to open a blank email. You can also prepopulate the To: field with the destination address by typing it after the colon, if you know it. To set the default email client on Windows, choose Apps > Default Apps from Settings; over on macOS, pick Mail > Settings > General from Apple Mail.

Run Instant Google Searches

On many Google searches, you get the answer above the list of links on the results page. These “instant” searches work in Chrome too. Ask about facts (like the height of the Eiffel Tower or the mass of Jupiter), celebrity ages, the days until a certain date, current stock prices for a company, the size of countries, the authors of books, and so on.

The original version of this story appeared in Quanta Magazine.

The bottom of the ocean is cold, dark, and under extreme pressure. It is not a place suited to the physiology of us surface dwellers: At the deepest point, the pressure of 36,200 feet of seawater is greater than the weight of an elephant on every square inch of your body. Yet Earth’s deepest places are home to life uniquely suited to these challenging conditions. Scientists have studied how the bodies of some large animals, such as anglerfish and blobfish, have adapted to withstand the pressure. But far less is known about how cells and molecules stand up to the squeezing, crushing weight of thousands of feet of seawater.

“The animals that live down in the deep sea are not ones that live in surface waters,” said Itay Budin, who studies the biochemistry of cell membranes at the University of California, San Diego. “They’re clearly biologically specialized. But we know very little, at the molecular level, about what is actually determining that specialization.”

In a recent study published in Science, researchers took the deepest look yet at how cells have adapted to life in the abyss. In 2018, Budin met Steve Haddock, a deep-sea biologist, and they combined forces to investigate whether cell membranes—specifically, the lipid molecules that membranes are made of—could help explain how animals have come to thrive in such a high-pressure environment.

To find out, they turned to comb jellies, the simple, diaphanous animals that Haddock studies at California’s Monterey Bay Aquarium Research Institute (MBARI). Led by his student Jacob Winnikoff, the interdisciplinary team discovered that the membranes of comb jellies that reside in the depths are made of lipid molecules with completely different shapes than those of their shallow-water counterparts. Three-quarters of the lipids in the deep-sea comb jellies were plasmalogens, a type of curved lipid that is rarer in surface animals. In the pressure of the deep sea, the curvy molecule conforms to the exact shape needed to support a sturdy yet dynamic cell membrane.

“It’s an amazing paper … with quite profound implications,” said Douglas Bartlett, who studies how microbes sustain life at depth and pressure at the University of California, San Diego and was not involved in the new study. “They provide another explanation for how the lipids of deep-sea animals, and likely deep-sea microbes and a range of organisms, are adapted in a way that’s pressure-specific.”

“They are looking into an area that, to a large degree, has not been explored,” said Sol Gruner, who researches molecular biophysics at Cornell University; he was consulted for the study but was not a co-author.

Plasmalogen lipids are also found in the human brain, and their role in deep-sea membranes could help explain aspects of cell signaling. More immediately, the research unveils a new way that life has adapted to the most extreme conditions of the deep ocean.

Insane in the Membrane

The cells of all life on Earth are encircled by fatty molecules known as lipids. If you put some lipids in a test tube and add water, they automatically line themselves up back to back: The lipids’ greasy, water-hating tails commingle to form an inner layer, and their water-loving heads arrange together to form the outer portions of a thin membrane. “It’s just like oil and water separating in a dish,” Winnikoff said. “It’s universal to lipids, and it’s what makes them work.”

For a cell, an outer lipid membrane serves as a physical barrier that, like the external wall of a house, provides structure and keeps a cell’s insides in. But the barrier can’t be too solid: It’s studded with proteins, which need some wiggle room to carry out their various cellular jobs, such as ferrying molecules across the membrane. And sometimes a cell membrane pinches off to release chemicals into the environment and then fuses back together again.

For a membrane to be healthy and functional, it must therefore be sturdy, fluid, and dynamic at the same time. “The membranes are balancing right on the edge of stability,” Winnikoff said. “Even though it has this really well-defined structure, all the individual molecules that make up the sheets on either side—they’re flowing around each other all the time. It’s actually a liquid crystal.”

One of the emergent properties of this structure, he said, is that the middle of the membrane is highly sensitive to both temperature and pressure—much more so than other biological molecules such as proteins, DNA or RNA. If you cool down a lipid membrane, for example, the molecules move more slowly, “and then eventually they’ll just lock together,” Winnikoff said, as when you put olive oil in the fridge. “Biologically, that’s generally a bad thing.” Metabolic processes halt; the membrane can even crack and leak its contents.

To avoid this, many cold-adapted animals have membranes composed of a blend of lipid molecules with slightly different structures to keep the liquid crystal flowing, even at low temperatures. Because high pressure also slows a membrane’s flow, many biologists assumed that deep-sea membranes were built the same way.

But it turns out these researchers weren’t getting the full picture. It would take an unexpected collaboration between biochemists and marine biologists, and more advanced technology, to see that deep-sea membranes had evolved a different way of going with the flow.

Going Deep

Comb jellies, or ctenophores, are voracious predators in fragile bodies. They are the largest animals that swim with cilia, which are lined up in rows known as combs, and they feed on a wide range of prey. Genetic evidence suggests that they were the first organisms to branch off the animal tree on their own evolutionary path. Though they resemble jellyfish in some ways, humans are actually more closely related to jellyfish than ctenophores are. And they have successfully colonized all kinds of ocean habitats, from surface waters to ocean trenches, and from the tropics to the poles.

You would expect such a wide-ranging group to be adaptable, and indeed comb jellies from the deep are built differently than those that live near the ocean’s surface. “You collect the deep guys, and you bring them up to the surface, and they just fall apart,” Bartlett said. “They just melt away. It’s really quite dramatic.” Similarly, if the ones adapted to shallow water end up at depth, they beat their cilia faster and faster, and eventually die. But no one really knew the molecular differences that separated them.

In 2018, Haddock, an expert on comb jellies, attended a conference on the origin of eukaryotes. After watching Budin present research on cell membranes’ response to temperature, he approached the lipid expert. Haddock had a graduate student, Winnikoff, who wanted to study adaptations to extreme pressure. It was known that lipids are sensitive to pressure, so cell membranes were a prime target for investigation. They decided to collaborate.

Haddock, Budin, and Winnikoff started by collecting comb jellies from different parts of the ocean. In scuba gear, Winnikoff carefully coaxed comb jellies from Monterey Bay’s surface waters into jars. From one of MBARI’s oceanographic vessels, he helped operate a deep-sea robot to collect comb jellies from depths of 12,000 feet. To control for the effects of the cold temperatures in the deep sea, he and Budin asked friends who were on their own expedition to gather surface comb jellies from frigid Arctic waters. In total, the team collected 66 animals from 17 related species.

By the time the molecular part of the project was set to begin, the pandemic had hit. So Winnikoff set up an experiment in his garage. Using a fluorescence spectrometer, he sent rays of ultraviolet light into test tubes filled with small globs of membrane material from the creatures they’d collected. The results puzzled him. The deep-sea membranes didn’t become more fluid as he raised the temperature—a response considered universal among lipid membranes.

So he and Budin consulted Gruner, the former director of Cornell’s particle accelerator. If they really wanted to know what was happening in the membranes, Gruner said, they would need powerful, high-energy X-rays. And he knew the perfect source.

Under Pressure

Buried 50 feet beneath the main athletic fields at Cornell is a synchrotron: a particle accelerator that uses a high-frequency electric field and a low-frequency magnetic field to speed up charged particles. Part of the facility, which Gruner fought to establish, may as well have been designed for studying deep-sea cell membranes. Its small-angle X-ray scattering operation, which opened in 2020, can not only distinguish the finer details and shapes of molecules such as lipids, but also increase and decrease the pressure they’re under.

The team experienced some pressure, too, as they had to endure late nights to make the most of their limited time at the facility. The powerful X-rays they shot at their lipid samples revealed the clearest picture yet of cell membranes from the abyss. The deep-sea comb jellies had membrane lipids that, at our standard atmospheric pressure, have a curvier shape than those in surface cell membranes. The animals had especially increased production of the group of lipids known as plasmalogens.

“In these deep-sea comb jellies, [plasmalogens] can make up three-quarters of all the lipids, and we’re talking about all the membrane lipids in the entire body of the animal, which is kind of crazy,” Winnikoff said. “We did a lot of checks to make sure that wasn’t a mistake.”

At the surface, a plasmalogen has a small phosphate head and a pair of wide, flaring tails, resembling a badminton shuttlecock, he said. But at high pressure, the tails squeeze together to form the necessary sturdy yet dynamic structure.

“They start their lipids at a different shape,” Budin said. “So when you compress them, they still maintain the right Goldilocks shape that you see in our own cells, but at these extreme pressures.” Budin and Winnikoff named this novel modification “homeocurvature adaptation.”

Taking a plasmalogen membrane to the deep sea is like pushing down on a spring, Bartlett said. At the surface, when the spring’s tension is released, it extends dramatically. “That’s when you can imagine the cells, their membranes, falling apart.” Meanwhile, if a surface membrane with straighter lipids is brought down to the deep, it compresses too much and becomes too rigid to function properly.

Notably, curvy plasmalogens were not present in comb jellies from the cold, shallow waters of the Arctic. “The composition of the membrane almost restricts the organisms to a particular pressure range,” said Peter Meikle, a lipid biologist who works on plasmalogens at the Baker Heart and Diabetes Institute in Australia and was not involved in the study.

But Budin wanted to see these lipids in action, and something occurred to him during a late session at the synchrotron. “In the middle of the night when you’re deliriously tired,” he said, sometimes you have a good idea. He stumbled on a paper with an intriguing approach to studying lipids. The authors had engineered Escherichia coli bacteria to produce plasmalogens in their membranes instead of their normal lipids. Budin realized that his team could similarly coax the bacteria to produce more plasmalogens and pressurize them to see how the membranes held up in living cells.

Following the paper’s methods, they showed that the bacteria with plasmalogen membranes could indeed better tolerate pressure than typical ones. These experimental membranes were made up of only 20 percent plasmalogens, but it was “enough to make a difference,” Winnikoff said.

Bartlett was impressed that the effect of the curved lipid shapes occurred in such unrelated species. “What is likely to come out of this is that we’ll find that this principle of homeocurvature adaptation will become a universal property of life,” he said.

Curvy Flexibility

Plasmalogens aren’t limited to the deep sea. They’re also found to varying degrees in other organisms, including humans. The percentage of plasmalogens within humans depends on the cell type. In the liver, plasmalogens make up 5 percent of phospholipids. In muscles, they can range between 20 percent and 40 percent. And in the brain, they make up about 60 percent.

In fact, the deterioration of plasmalogens has been linked to neurodegenerative disorders such as Alzheimer’s disease. “The evidence suggests that the plasmalogens are more protective,” said Meikle, who studies plasmalogens because of their links to mammalian health.